Return
Graphene-integrated hybrid Au-Ag plasmonic terahertz metasurface with Fano resonances and XGBoost-assisted performance prediction for high-sensitivity label-free peptide biosensing
V
G
K
A
U
DOI:10.1016/j.diamond.2026.114051.png)
Abstract
En 中文
Peptide biomarkers play a crucial role in disease diagnosis and therapeutic monitoring, creating a growing demand for highly sensitive, rapid, and label-free detection platforms. To address the limitations of conventional biochemical sensing methods, this study proposes a hybrid graphene–Au–Ag plasmonic terahertz metasurface biosensor incorporating a Cu T-shaped resonator, an Au circular ring, and an Ag square-ring coupler integrated on a WS₂ substrate. The nested resonator architecture generates electromagnetically induced transparency-like Fano resonances and strong localized plasmonic hotspots, enabling enhanced light–matter interaction for peptide sensing. Comprehensive full-wave simulations were conducted using COMSOL Multiphysics under four material distribution configurations and varying graphene chemical potentials. The optimized configuration demonstrated stable resonance characteristics up to an incidence angle of 80° and achieved a maximum sensitivity of 1150 GHz/RIU, a figure of merit of 4.399 RIU−1, and a detection limit of 0.416 across peptide refractive indices ranging from 1.606 to 1.706 RIU. Electric-field analysis revealed pronounced confinement at 0.98 THz, corresponding to the strongest sensing response. Furthermore, XGBoost-based prediction achieved R2 values exceeding 0.9997 and accuracies above 99.92%. These findings establish the proposed metasurface as a promising platform for future portable, high-performance terahertz biosensing and intelligent peptide diagnostics.
Keywords:
THz biosensor
Fano resonance
Peptide detection
XGBoost machine learning
Journal
IF:
5.1
Papers:
2.1K
Citations:
2.4W
